Summary

A Japanese-language book on context engineering for large language models (LLMs), by Hirosato Gamo, a cloud solutions architect and evangelist at a foreign-owned IT company. It defines context as the varied input given to an LLM, including the prompt, and treats context engineering as the body of techniques for deciding what to give, what to discard, and how to keep the limited input space in good condition. The book explains the basic mechanisms of AI models, starting from how the Transformer works, and the behavior of APIs from the viewpoint of context. It covers instruction prompt design, preparing data sources for RAG (Retrieval-Augmented Generation), and automating workflows with AI agents. Taking LLM quirks such as Lost in the Middle and the volatility of conversation history into account, it presents a view of designing context without relying only on prompt tweaks.

Target Readers

• IT engineers interested in developing LLM applications who want to learn how the Transformer and the APIs handle context
• Developers working to improve the accuracy of a RAG system or an AI agent who want to learn how to prepare data sources and design context
• Engineers who work on developing LLM applications and want to learn how instruction prompts, API behavior, RAG, and AI agents relate to context

Colophon

Publisher

Gijutsu-Hyohron

ISBN

978-4-297-15419-6

Published

Feb 2026

List price

¥3,300incl. taxMay differ from the actual selling price on Amazon